-
-
Notifications
You must be signed in to change notification settings - Fork 162
System monitor : hardware support
Matthieu Houdebine edited this page Sep 23, 2023
·
24 revisions
Computer hardware support for system monitor depends on the HW manufacturer and the OS.
On Windows platforms, you have 2 ways of getting hardware sensors:
- LibreHardwareMonitor which offers optimal HW support, but requires administrator rights
- Python libraries, with some HW limitations (see below)
On other platforms (Linux, macOS) all HW sensors data will be fetched from Python libraries.
Check this table to know if your hardware is supported by system monitor on your OS:
✅ Supported / ❕ Supported with limitations (see notes below) / ❌ Unsupported / ❔ Untested
| Windows (LibreHardwareMonitor integration) | Windows (Python libraries) | Linux | macOS | Raspberry Pi ❕ | |
|---|---|---|---|---|---|
| CPU | ✅ Intel ✅ AMD |
✅ Intel ✅ AMD ❌ No CPU temperature |
✅ Intel ✅ AMD |
✅ Intel ✅ AMD ❔ CPU temperature |
✅ ARM |
| GPU | ✅ Nvidia ❕ AMD ✅ Intel ❕ Multiple GPUs |
✅ Nvidia ❕ AMD (❌ not with Python 3.12+) ❕ Multiple GPUs |
✅ Nvidia ❕ AMD (see note for Python 3.12+) ❕ Multiple GPUs |
✅ Nvidia ❕ AMD (see note for Python 3.12+) ❕ Multiple GPUs |
❌ |
| GPU FPS | ❌ Nvidia ✅ AMD ❌ Intel |
❌ | ❌ | ❌ | ❌ |
| RAM | ✅ | ✅ | ✅ | ✅ | ❔ |
| Disk | ✅ Only system disk | ✅ Only system disk | ✅ Only system disk | ✅ Only system disk | ❔ |
| Network | ✅ | ✅ | ✅ | ✅ | ✅ |
Notes:
-
Raspberry Pi 0 / 1 / 2 / 3:
- If the display refresh is too slow, follow this procedure.
-
Multiple GPUs / all platforms:
- You will not be able to display stats from several GPUs: the System Monitor will auto-select the best GPU to use for "GPU" stats
-
AMD GPU / Linux & MacOS:
- To get AMD GPU support on Linux or MacOS with Python 3.12+, follow these instructions.
-
AMD GPU / Windows:
- AMD GPUs not yet supported with Python libraries and Python 3.12+: use LibreHardwareMonitor or downgrade your Python version
- GPU memory usage not available (this is a hardware limitation: AMD GPUs do not offer a public API for this).
- Use LibreHardwareMonitor integration if you have a 2019+ GPU, it is not supported by pyadl Python library.
- Nvidia GPU / all platforms: GPU usage may be unsupported if this is not your main GPU (e.g. on a laptop with an integrated Intel GPU)
Feel free to add additional info to this table as a PR if you test on other OS / hardware.